Product and Offer Design Studio
Outcome
Help product managers design commercially sound mobile and broadband offers by testing candidate bundles against segment usage, margin, and regulatory constraints before pricing committee approval and launch.
| Portfolio attribute | Value |
|---|---|
| Primary persona | Product manager |
| Primary aggregate | OfferDesign |
| Cosmos DB containers | 3 |
| Grounding data | offers, segments, usage, campaigns, policies |
| Development estimate | USD 60-220 monthly dev estimate |
| Runtime options | Python FastAPI and C# ASP.NET Core |
Productivity hypothesis
These measures reflect the workflow, evidence, decision boundary, and industry risk for this use case.
| Impact type | Measure | Conservative internal assumption | Customer measurement |
|---|---|---|---|
| Productivity | Design brief cycle time | 10-18% reduction | (baseline median design-cycle minutes - assisted median minutes) / baseline median minutes * 100 |
| Quality or risk | First-pass acceptance | 5-10 percentage-point improvement | assisted percentage accepted without material rework - baseline percentage accepted without material rework |
Important: Conservative internal pilot assumptions only; not public benchmarks, commitments, or expected customer results. Validate against a comparable baseline or matched control cohort with mandatory human review.
Rationale: Offers, segments, usage, campaigns, and policies support faster product-concept iteration, although approvals may dominate elapsed time.
Pilot guardrail: Require policy, pricing, compliance, campaign, and commercial approval before launch.
Customer validation plan
- Baseline the current workflow with representative users and tasks.
- Run the accelerator on the same task mix with human review retained.
- Measure design brief cycle time and first-pass acceptance with the documented formulas.
- Replace internal assumptions with customer-observed results before a business case.
Why Azure Cosmos DB
This AI solution needs an operational data plane, not only a vector index:
- Hybrid vector and full-text retrieval combines semantic recall with exact identifiers, codes, names, and domain language.
- Hierarchical partition keys and metadata predicates keep tenant and domain boundaries in the normal query path.
- Operational JSON, source chunks, workflow and agent state, feedback, and audit records remain together for low-latency reads and governed writes.
- Change feed reacts to source and operational updates without polling.
- Autoscale throughput, configurable consistency, and multi-region distribution support variable demand and enterprise availability.
Domain workflow

| Stage | Actor | Input → output | Approval |
|---|---|---|---|
| 1. Set the commercial objective and target segments | Product manager | Growth, retention, or ARPU objective with segments → OfferDesign shell with target segments | Policy controlled |
| 2. Compose candidate bundles from catalogue components | Design agent | Component catalogue and segment usage profiles → Candidate bundle variants with cost structure | Policy controlled |
| 3. Simulate take-rate, margin, and cannibalization | Financial modelling service | Candidate bundles and segment elasticity → MarginSimulation set with sensitivity ranges | Policy controlled |
| 4. Screen regulatory and technical feasibility | Compliance and feasibility agent | Bundle structure and market constraints → Blocking and advisory findings per market | Policy controlled |
| 5. Pricing committee approval | Pricing committee | Shortlisted offers with simulations and findings → Approved offer released to configuration and launch | Human required |
Technical architecture

Cosmos DB data model
| Container | Hierarchical partition key | Documents | Search paths |
|---|---|---|---|
offerDesigns | /tenantId, /productPortfolioId | OfferDesign, SegmentUsageProfile, MarginSimulation | /objectiveVectorEmbedding, /objectiveVector |
offerComponentCatalog | /tenantId, /componentFamilyId | OfferComponentCatalogItem | /componentDescriptionVectorEmbedding, /componentDescriptionVector |
telecomRegulationLibrary | /tenantId, /marketCode | RegulatoryConstraint | /constraintVectorEmbedding, /constraintVector |
Query contract
| Query | Container | Scope | Purpose |
|---|---|---|---|
| Design workspace assembly | offerDesigns | single-partition | Read the offer, its simulations, and target segment profiles in one portfolio partition query. |
| Margin threshold screen | offerDesigns | single-partition | Keep only candidate offers whose projected margin clears the portfolio floor. |
| Market-applicable constraints | telecomRegulationLibrary | cross-partition-bounded | Knowledge-container query against telecomRegulationLibrary (bounded fan-out across the /tenantId + /marketCode partitions named by the shortlist): retrieve the regulatory rules in force for the launch markets, fanned out across the small fixed list of target market partitions. |
| High-churn segment targeting | offerDesigns | single-partition | Prioritize segments where churn and ARPU justify a retention offer. |
Python and C# services route each document type to its declared container and hierarchical partition key. ETag-based writes protect mutable aggregates, while append-only records and bounded cross-partition queries preserve scale.
Azure deployment

Clone the source repository, authenticate with Azure, and run:
Set-Location accelerators/telecommunications/product-and-offer-design-studio
az login
./Deploy.ps1 -Runtime python -Location eastus2
Change python to csharp to deploy the .NET implementation.
Explore the implementation
- Launch signed-in demo
- Implementation article
- Domain data model
- Workflow and API contract
- Python application
- C# application
- Domain container Bicep
- Platform Bicep
- Cost and configuration
- Source repository